Object-Oriented Information Extraction of Farmland Shelterbelts from Remote Sensing Image

Abstract

It has become an important means of shelterbelts surveying using high resolution remote sensing image to access the distribution of farmland shelterbelts. However, traditional classifications of remote sensing image based on spectrum characteristics of single pixel, and didn’t consider the factors including relativity and structure characteristics of the neighboring pixels, which will lead to lower accuracy of feature extraction for high resolution remote sensing image. On the basis of object-oriented classification method and the module of ENVI Feature Extraction, the paper extracted the shelterbelts distribution through image segmentation and rules establishment for the Spot5 high resolution remote sensing image in the Midwest of Jilin Province, and the extraction accuracy is 91.3%.The result shows that the method can accurately extract farmland shelterbelts from high resolution remote sensing image.